INTERNATIONAL JOURNAL OF SYSTEMATIC
OO20-7713/82/01Oo77-05$02.OO/O
BACTERIOLOGY, Jan.
1982, p. 77-81
Vol. 32, No. 1
Clostridium perenne and Clostridium paraperfringens: Later
Subjective Synonyms of Clostridium bara ti
ELIZABETH P. CATO, LILLIAN V. HOLDEMAN,
AND
W. E. C. MOORE
Department of Anaerobic Microbiology, Virginia Polytechnic Institute and State University, Blacksburg,
Virginia 24061
Identical electrophoretic patterns of cellular proteins were detected with ATCC
27638, ATCC 25782, and ATCC 27639, the type strains of Clostridium barati
(Prevot 1938) Holdeman and Moore 1970, C. perenne (PrCvot 1940) McClung and
McCoy 1957, and C. paraperfringens Nakamura et al. 1970, respectively, all of
which are cited in the Approved Lists of Bacterial Names. The morphological and
biochemical reactions of these strains and of other reference strains of these
species were also similar. On these bases, it is herein proposed that these three
names are synonyms. According to the rules of the Bacteriological Code, in those
cases in which names on the Approved Lists compete for priority, the priority is
determined by the date of the original publication of the name before 1 January
1980. According to our interpretation of this rule, the specific epithet “barati” has
priority. The correct name of this organism, then, is Clostridium barati, and C .
perenne and C . paraperfringens are later subjective synonyms.
During examinations of clostridial cellular
proteins that were detected by polyacrylamide
slab gel electrophoresis (7), we found that protein patterns produced from the type strains of
C. b a r d (Prevot 1938) Holdeman and Moore
1970, C.paraperfringens Nakamura et al. 1970,
and C. perenne (PrCvot 1940) McClung and
McCoy 1957 were identical. We therefore reexamined the phenotypic characteristics and the
electrophoretic patterns of other strains in our
collection that had been identified as members
of these three species by the investigators who
first described them.
RESULTS AND DISCUSSION
Electrophoretic patterns of soluble proteins
from cells in 6-h-old cultures of the strains
examined are shown in Fig. 1. Patterns from two
phenotypically similar species of saccharolytic
clostridia, C. absonum and C . perfringens, are
included.
Biochemical properties of strains of C. barati,
C . perenne, and C . paraperfringens were found
to be nearly as similar as the electrophoretic
patterns and agreed in most respects with the
properties given in the original descriptions. The
discrepancies and variations that occurred are
listed in Table 2. Although all the strains in this
study reduced nitrate, this characteristic has
been found to be variable among strains (3, 8).
Prevot described Injlabilis barati in 1938 (10)
and assigned it to his new genus Injlabilis because it had central to subterminal spores and
was not motile. The species was transferred to
the genus Clostridium by Holdeman and Moore
in 1970 (4). In 1940, Prevot described Acuformis
perennis, a saccharolytic, nonmotile species that
formed terminal spores (11). This species was
placed in the genus Clostridium by McClung and
McCoy in 1957 (6). We frequently had difficulty
differentiating between these two species because the reported major difference was the
location of the spores. Most strains of these
species, including the labeled strains received
from PrCvot, sporulate very poorly, but all
strains survive heating at 80°C for 10 min. The
appearance and location of spores depend not
only on the medium used but also on the length
MATERIALS AND METHODS
Bacterial strains. The strains studied and their
sources and original designations are listed in Table 1.
Those with only a VPI number had been identified in
our laboratory by conventional phenotypic tests (3).
Electrophoresis. The electrophoretic method described by Moore et al. (7) was used to compare
cellular proteins released from the bacterial cells.
Organisms were incubated for 6 h in 5 ml of brain heart
infusion broth supplemented with 0.1% calcium carbonate.
Culture media and methods. For determinations of
the physiological characteristics of the strains, prereduced, anaerobically sterilized media and anaerobic
methods described in the Virginia Polytechnic Institute and State University Anaerobe Laboratoty Manual (3) were used. Hydrogen sulfide production was
tested in SIM (sulfide-indole-motility) medium (BBL
Microbiology Systems, Cockeysville, Md.).
77
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sun, 18 Jun 2017 21:03:27
78
CATO, HOLDEMAN, AND MOORE
INT. J . SYST.BACTERIOL.
TABLE 1. Bacterial strains examined
VPI
no.
ATCC
no.
Original
identification
444 3
26738a
lnflabilis barati
4624
Inflabilk barati
4475
25782a
Acuformk perennis
(C. perenne)
6921
Source and other numbers
A. R. Prevotb no, 2227, canine renal epithelioma
A. R. Prevot no. 1236, boil
A. R . Prevot no. 11 16D, hog h e r
Tucson Medical Center, boil on thigh
8204
Florida State Health Dept., peritoneal fluid
9436
VA Hospital, Sepulveda, CA
California State Health Dept., ear
9919
6907A
27639a
6908
6909A
26740
4846
C . paraperfingens
S. Nakamurac no. G , gas gangrene
C. paraperfringens
S. Nakamura no. F1S 5
C. paraperfringens
S. Nakamura no. 3-3, normal human intestine
lnflabilis lacustris
A. R. Prevot no. 12/16B
6903A
27635
C. absonum
S. Nakamura no. HA9103, soil
13181
2755Sa
C. absonum
S. Nakamura no. HA7103, soil
5694
13124a
C. perfringens
Wellcome Lab no. CN 1491
aType strain (1 5)
bPasteur Institute, Paris, France
‘Kanazawa University, Kanazawa, Japan
of incubation, and they can vary markedly in the
same preparation. The photograph accompanying the original description of C . perenne (11)
shows spores that are definitely subterminal as
well as some that might be considered terminal.
In 1970, Nakamura et al. (9) published a
description of strains they had isolated from a
case of gas gangrene, from normal human intestinal contents, and from soil. They named the
species C . paraperfringens but did not designate
a type strain. In 1973, in an expanded study of
the properties of these strains (8) and in comparison with other saccharolytic strains of clostridia, they designated a type strain of C . paraperfringens and reported that this strain was
genetically homologous by deoxyribonucleic
acid (DNA)-DNA hybridization with reference
DNA prepared from strain 2227 of I . b a r d
Prevot (which is the type strain of C . barati
[Prevot] Holdeman and Moore 1970 [IS]) that
they had received from the Pasteur Institute in
Paris, France.
It is not clear why Nakamura et al. continued
to use the name C . paraperfringens for the
species. Prevot’s description of I . barati was
validly published (lo), the species is considered
legitimate (2), and the name C . barati appeared
on the 1980 Approved Lists of Bacterial Names
(1 5 ) .
Perhaps Nakamura et al. did not recognize I .
barati because the two labeled strains of I .
barati that they studied did not have the characters reported for “le bacille de Barat” by Tissier
(16). In citations of new species, Prevot often
attributed the name to previous authors even
though they had not named the species according to accepted rules. Thus, Prevot (10) cited the
species as “ I . b a r d (Tissier) nv. dnm.” Because the species was named by Prevot, not
Tissier, whether or not C . barati (PrCvot)Holdeman and Moore is the same organism as that
seen by Barat and Tissier, later described by
Tissier (16) as “le bacille de Barat,” has no
bearing on the validity of this species.
Perhaps Nakamura et al. placed PrCvot’s
strains of C. b a r d in their new species C .
paraperfringens on the basis of the discrepancies from PrCvot’s description rather than on
their first-hand analysis of authentic strains.
Some taxonomists take the position that the
“description” represents (i.e., “is”) the species
(E. F. Lessel, personal communication). Thus,
the type strain (description) would remain the
same, even if variation occurred in the culture as
a result of continued preservation. But this
position does not allow for errors in the descriptions that are based on erroneous or incomplete
analytical results or are caused by technical
errors, misprints, oversights, or limitations of
methods. Most taxonomists take the position
that the species is based on the type strain and
that the description can be modified on the basis
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sun, 18 Jun 2017 21:03:27
SYNONYMS OF CLOSTRIDIUM BARATI
VOL. 32, 1982
mentation of milk, and fermentation products)
can be explained. When Prevot described a
species, he often chose to emend any previously
published description by other authors rather
than amend it. Since he believed his strains were
similar to “le bacille de Barat” of Tissier (16),
PrCvot included in his description of I . barati
some of the characteristics Tissier had listed.
Prevot did not include motility (as described by
Tissier) as a characteristic (his genus Znflabilis
included only nonmotile species), probably because he was aware that in early work Brownian
movement often was mistaken for motility. Although Tissier stated that the organism produced only acetic acid, Pr6vot did not report the
composition of the acid products (as formic,
butyric, and lactic) until 1948 (12). He believed
that Duclaux distillation and earlier procedures
were inaccurate for identifying mixtures of
short-chained fatty acids (A. R. Prevot, personal communication). As detected by gas chromatographic analysis, the strains produced formic, acetic, butyric, and lactic acids, with traces
of propionic acid. The stormy fermentation of
milk was irregularly observed. We reported (3)
that the reaction in milk is variable (see also
Table 2 footnote and Table 3).
Almost certainly, Prevot included mannitol
fermentation as a property of the species because Tissier stated that mannitol was fermented. We have no indication that PrCvot ever
detected fermentation of mannitol by any culture he recognized as I . barati. On our copies of
his original work sheets for two of the strains, no
mention is made of mannitol fermentation by
Prkvot strain 2227, and the record for Prevot
strain 1236 bears the notation that “all except
mannitol” (translated) were fermented. We ex-
Fig. 1. Electrophoretic protein patterns of
Clostridium isolates. Lane ( 1) reference strain
U4-20 of Streptococcus faecalis; ( 2 , 3 ) C. barati
4443, 4624; (4) C. paraperfringens 6908; ( 5 , 6 ,
7, 8, 9 ) C. perenne 4475, 6921, 8204, 9436,
9919; (10, 1 1 , 12) C. paraperfn‘ngens 6907A,
6909A, 4846; (13, 14) C. absonum 13181,
6903A; (1 5 ) C. perfringens 5694.
of improved analyses of the type strain. If the
strain really does change during storage, this too
is a valid property of the species.
The discrepancies between Prevot’s description and the reactions of his strains examined by
Nakamura et al. and by us (lipolytic activity,
fermentation of mannitol and inulin, stormy fer-
TABLE 2 . Reactions of strains of Clostridiuma
Reaction
C. barati
e;evorb
4624
Prgvotc
-
-
+
w
-
w
-
+
-
-
-
+
+
+
+
+
-
Acid from
Amygdalin
Arabinose
Esculin
Glycogen
Inulin
Mannitol
Ribose
Starch
NOj reduced
-
Lipase
+
H2S
+
+
+
C. paraperfringens
C. perenne
4443
6921
-
-
-
-
-
-
+
4475
-
+
w
+
-
-
-
-
79
8204 9436 9919
_
_
_
-
_
+
_
_
w
-
_
-
-
_
_
-
-
w
w
_
-
6908
6909A
4846
-
-
w
-
-
w
w
w
-
-
w
-
w
-
-
-
-
-
-
V
w
w
-
+
+
+
-
-
_
+
_
6907A
-
-
+
-
+
+
-
-
-
-
-
-
-
-
+
-
-
-
-
+
-
-
+
-
+
found that all of the strains produced acid from cellobiose, fructose, galactose, lactose, maltose, mannose, salicin. and sucrose. None
produced acid from erythritol, inositol. melezitose, melibiose. raffinose, rhamnose, sqrbitol, trehalose, or xylose. All strains hydrolyzed esculin.
Neither gelatin, milk, nor meat was digested. An acid curd was formed in milk, Indole was not formed. Lecithinase was produced by all strains;
none produced catalase or urease. Large amounts of hydrogen gas were produced. N o strain was motile; none was toxic to mice.
Symbols: -, negative reaction; w, pH between 5.5 and 6 . 0 ; +, positive reaction or (carbohydrate cultures) pH below 5.5; v, reaction variable.
Where two reactions are given, the first was rhe more usual, the other observed less frequently.
c(ll) d ( 8 )
’(10. 12)
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sun, 18 Jun 2017 21:03:27
80
CATO, HOLDEMAN, AND MOORE
INT. J .
SYST.
BACTERIOL.
TABLE 3 . Results of Babcock butterfat analyses of 8-day-old cultures in homogenized
and whole milk
Inoculum
Sterile homogenized cow’s milk
Appearance
% Butterfat
Sterile whole goat’s milk
Appearance
% Butterfat
None
N o change
2.8
Semisolid precipitate
6.1
4475 C. perenne
Solid, smooth curd, gas
2.1
Hard shrunken curd
6.0
6907A C. paraperfringens N o change
2.0
Hard shrunken curd
6.2
N o change
2.0
Hard shrunken curd
5.5
4443 C. barati
amined four of the eight strains he isolated (14),
and none lowered the pH of mannitol medium.
The fermentation of mannitol and inulin reported by Tissier and listed by PrCvot (13) but not
found by Prevot, by Nakamura et al., or by us,
can be ascribed to contamination of those sugars
with glucose, as was later realized, or to reduction of the pH indicator often used to determine
acid production.
Prevot’s description of I . barati as lipidolytic
also was probably based on Tissier’s description. There is no indication on our copies of
Prevot’s records that he ever tested these strains
for this property. Tissier reported (16) that “in
milk containing 41 grams of ‘butter’ per liter,
only 26 grams were obtained after 8 days of
culture” (translated). The value of this observation is questionable. Before the Babcock test
was developed, a standard method for analysis
of milk quality (“richness”) was to determine
how much butter it would yield. Although there
was an apparent 36% loss of butter, butter itself
is only 60 to 70% butterfat. The remainder is not
fat but solids (including lecithin and protein) and
water. Furthermore, butterfat is lost by natural
degradation if milk is held for 8 days. Without a
complete description of Tissier’s analysis for
butter, it is difficult to postulate how accurate
determinations of the recovery of butter could
be made from the incubated milk which formed
dense solid curds (stormy fermentation).
Nevertheless, we ran Babcock butterfat (as
opposed to butter) analyses (1) on samples of
sterilized homogenized cow’s milk and whole
goat’s milk that was inoculated and incubated
for 8 days. It was necessary to break mechanically the protein clots present in some cultures.
Even so, the remaining small particles probably
caused analytical errors. The averages of triplicate analyses are reported in Table 3. These
observations indicate that butterfat (and probably butter yield) may decrease after 8 days of
incubation.
Nakamura et al. might have questioned
whether the strains received from the Pasteur
Institute were authentic. However, the characteristics of the strains are those observed by
PrCvot, who originally identified the strains and
is author of the species. Therefore, there should
be no question about the authenticity of the
strains. On the basis of these strains, including
the type strain, and the information given above,
the description of C. barati was amended by
Holdeman and Moore in 1970 (4) to agree with
the biochemical reactions of the strains.
From our analyses of the type strains, and the
DNA-DNA homology data of Nakamura et al.,
it is clear that C. paraperfringens Nakamura,
Tamai, and Nishida 1970, is a synonym of Clostridium barati (Prevot) Holdeman and Moore
1970, both of which appear in the 1980 Approved Lists of Bacterial Names.
According to Rule 24b of the International
Code of Nomenclature of Bacteria (5): “If two
names compete for priority and if both names
date from 1 January 1980 on an approved list,
the priority shall be determined by the date of
the original publication of the name before 1
January 1980.’’
Because strains (including, in each case, the
type strain) of C. paraperfringens are homologous with those of C. barati (8) and the strains of
C. barati and C . perenne have common morphological and biochemical properties and identical
electrophoretic patterns of soluble cellular proteins which correlate well with DNA homology
analyses, we propose that the names C. paraperfringens and C . perenne be rejected as later
synonyms and that all of these strains be included in the species C. barati (Prevot 1938) Holdeman and Moore (1970). The type strain of this
species is ATCC 27638 (+- VPI 4443 + Pasteur
Institute 2227).
Rule 24b does not indicate clearly whether
priority is to be based on the binary name (as
given in the 1980 List) or on the specific epithet.
However, Rule 23a (Note 1) and Rule 32b clearly indicate that, in cases of priority, the specific
epithet (i) is treated independently, and (ii) is not
rendered illegitimate by publication as part of a
binary combination in which the generic name
(of the combination) is illegitimate. Thus, Influbilis barati Prevot 1938, which was (but is not
now) a legitimate binary name, would establish
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sun, 18 Jun 2017 21:03:27
SYNONYMS OF CLOSTRIDIUM BARATI
VOL.32, 1982
priority of the specific epithet “barati” over that
in Acuformis perennis Prevot 1940 and over
those in any other subsequent, related binary
combinations containing the currently accepted
generic name Clostridium:e.g., C . perenne (Prevot) McClung and McCoy 1957 and C. paraperfringens Nakamura et al. 1970. For these reasons we regard the specific epithet “barati” as
having priority for the species name, even
though the basonym, “Injabifis barati,” is not
on the Approved Lists of Bacterial Names.
ACKNOWLEDGMENTS
We gratefully acknowledge the assistance of Donald E.
Hash in the electrophoretic analyses and of Luba Fabrycky in
the bacteriological analyses.
This work was supported by Public Health Service grant A1
15244-01A from the National Institute of Allergy and Infectious Disease.
REPRINT REQUESTS
Address reprint requests to: Elizabeth P. Cato, Department
of Anaerobic Microbiology, Virginia Polytechnic Institute and
State University, Blacksburg, VA 24061.
LITERATURE CITED
American Public Health Association. 1960. Babcock method-official, p. 373-375. I n Standard methods for the
examination of dairy products, 11th ed. American Public
Health Association, Inc., New York.
Buchanan, R. E., J. G. Holt, and E. F. Lessel, Jr. (ed.).
1966. Index bergeyana, p. 614. The Williams & Wilkins
Co., Baltimore.
Holdeman, L. V., E. P. Cato, and W. E. C. Moore (ed.).
1977. Anaerobe laboratory manual, 4th ed. Virginia Polytechnic Institute and State University, Blacksburg.
81
4. Holdeman, L. V., and W. E. C. Moore. 1970. Clostridium,
p. 60. In L. V. Holdeman and W. E. C. Moore (ed.),
Outline of clinical methods in anaerobic bacteriology, 2nd
rev. Virginia Polytechnic Institute, Blacksburg.
5. Lapage, S. P., P. H. A. Sneath, E. F. Lessel, V. B. D.
Skerman, H. P. R. Seeliger, and W. A. Clark (ed.). 1975.
International code of nomenclature of bacteria. American
Society for Microbiology, Washington, D.C.
6. McClung, L. S., and E. McCoy. 1957. Genus I1 Clostridiurn Prazrnowski 1880, p. 673. In Breed, Murray, and
Smith (ed.), Bergey’s manual of determinative bacteriology, 7th ed. The Williams & Wilkins Co., Baltimore.
7. Moore, W. E. C., D. E. Hash, L. V. Holdeman, and E. P.
Cato. 1980. Polyacrylamide slab gel electrophoresis of
soluble proteins for studies of bacterial floras. Appl.
Environ. Microbiol. 39:900-907.
8. Nakamura, S., T. Shimamura, M. Hayase, and S. Nishida.
1973. Numerical taxonomy of saccharolytic clostridia,
particularly Clostridium perfringens-like strains: descriptions of Clostridium absonum sp. n. and Clostridium
paraperfringens. Int. J. Syst. Bacteriol. 23:419-429.
9. Nakamura, S., K. Tamai, and S. Nishida. 1970. Criteria for
identification of Clostridium perfringens. 6. Clostridium
paraperfringens sp. nov. Med. Biol. 80137-140.
10. Prbvot, A. R. 1938. Etudes de systkmatique bacterienne.
IV. Critique de la conception actuelle du genre Clostridiurn. Ann. Inst. Pasteur Paris 61:77.
11. Prevot, A. R. 1940. Recherches sur la flore anaerobie de
l’intestin humain: Acuformis perennis nov. sp. C.R. Seances SOC.Biol. Paris 133574-577.
12. Prevot, A. R. 1948. Manuel de classification et de determination des bacteries anakrobies. Masson and Co., Paris.
13. Prbvot, A. R. 1957. Manuel de classification et de determination des bacteries anakrobies, 3rd ed. Masson and Co.,
Paris.
14. Prbvot, A. R., A. Turpin, and P. Kaiser. 1967. Les batteries anaerobies. Dunod, Paris.
15. Skerman, V. B. D., V. McGowan, and P. H. A. Sneath.
1980. Approved lists of bacterial names. Int. J. Syst.
Bacteriol. 30:225-420.
16. Tissier, H. 1918. Le bacille de Barat. C.R. Seances SOC.
Biol. Paris 81:426-427.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sun, 18 Jun 2017 21:03:27